International Journal of Hydrogen Energy 32 (2007) 4978 – 4987

www.elsevier.com/locate/ijhydene

Performance simulation of metal hydride hydrogen storage device

with embedded filters and heat exchanger tubes
G. Mohan, M. Prakash Maiya, S. Srinivasa Murthy ∗
Refrigeration and Air Conditioning Laboratory, Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai 600 036, India

Received 6 June 2007; received in revised form 7 August 2007; accepted 8 August 2007
Available online 29 September 2007

Abstract
A practical metal hydride based hydrogen storage device would consist of many filters to distribute hydrogen gas and heat exchanger tubes
to cool or heat the hydride bed depending on whether hydrogen is being absorbed or desorbed. This paper presents the simulation of such a
device with LaNi5 as the hydriding alloy. A study of the geometric and operating parameters has been carried out to identify their influence
in the hydriding performance of the storage device.
䉷 2007 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.

Keywords: LaNi5 ; Metal hydride; Heat and mass transfer; Simulation; Hydrogen storage

1. Introduction Majority of the work on theoretical issues of metal hydride

in the recent past dealt with one dimensional treatment of metal
Metal hydrides as hydrogen storage media offer various ben- hydride beds. Ram Gopal and Srinivasa Murthy [5] developed a
eficial features such as high volumetric storage capacity, fast one dimensional heat transfer model based on the experiments
reaction kinetics, large number of charge–discharge cycles and carried out on a cylindrical metal hydride reactor filled with
safety. Metal hydrides are formed on exposure of certain met- MmNi4.5 Al0.5 . Their model and experiments confirmed the im-
als or alloys to hydrogen at ordinary temperatures and pres- portance of heat transfer as a rate controlling phenomena. Sev-
sures. One of the major drawbacks inherent to metal hydride is eral two dimensional models [6–11] treated the metal hydride
the poor heat and mass transfer characteristics, which call for bed as an exothermic porous media. These studies showed that
special attention in the design of the storage device. the hydrogen absorption rate is high near the reactor walls and
Early heat and mass transfer studies on metal hydrides were established the importance of bed thickness as the major geo-
based on the mass and energy balance of hydride beds (Powers metric parameter that influences the hydriding process. How-
and Cummings [1] and Yu et al. [2]). Corresponding research ever, the importance of physical configuration of the storage
activities on the kinetics of hydrogen sorption were carried out device and the associated geometric parameters are not fully
towards the advancement of metal hydrides as storage media. represented in the studies reported so far. The region near the
Goudy et al. [3] reported an experimental study on the reac- walls where the temperatures are lower gets saturated with hy-
tion kinetics of metal hydrides which ascertained the impor- drogen earlier and forms a reaction front which propagates into
tance of heat transfer in controlling the hydrogen evolution. the inner hotter regions of the reactor. While the mechanism of
Later on, Kawamura et al. [4] developed a lumped parame- hydride formation is independent of the geometry of the reac-
ter model for hydride beds. Since then, many models dealing tor, the shape of the reaction front and the speed of its move-
with the theoretical aspects of metal hydrides have appeared in ment depend on the reactor configuration.
literature. In general, configuration of metal hydride devices falls into
following categories:
∗ Corresponding author. Fax: +91 44 2257 4652. (i) A cylindrical tube with axial filter and outer cooling jacket
E-mail address: [email protected] (S. Srinivasa Murthy). filled with metal hydride in the annulus.
0360-3199/$ - see front matter 䉷 2007 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2007.08.007
 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987 4979

Nomenclature Greek letters

b bed thickness, m H 0 heat of formation, J kg−1
Cp specific heat, J kg−1 K −1  porosity
D diffusion coefficient, m2 s−1  permeability, m2
d diameter, m  dynamic viscosity, kg m−1 s−1
Ea activation energy, J mol−1  polar coordinate, radian
h heat transfer coefficient, W m−2 K −1  density, kg m−3
k thermal conductivity, W m−1 K −1
ṁ hydrogen mass absorbed, kg m−3 s−1 Subscripts
N number of cooling tubes amb ambient
p pressure, Pa e effective
r radial coordinate, m eq equilibrium
R universal gas constant f cooling fluid
(8.314 J mol−1 K −1 ) g gas
s pitch distance, m in inlet
t time, s o initial
T temperature, K s solid
u velocity, m s−1 sat saturated

(ii) Multiple heat transfer tubes and filters embedded in the areas of the alloy. Each tube could be identified as a nucleus
metal hydride bed in a cylindrical container. for the reaction fronts to propagate into the unsaturated region.
When all reaction fronts meet one another at the same instant,
The single tube design has been widely studied in literature the bed is left with the least amount of unsaturated alloy. Such a
referred above, mainly to understand the heat and mass transfer configuration could be considered as the most efficient in terms
mechanisms. Even though this can yield good heat and mass of hydriding performance.
transfer performance, it is poor from the overall system weight The reactor considered in the present study has an outer
point of view. For mobile applications requiring large hydro- container of radius ‘r1 ’, in which cooling tubes of radius ‘r2 ’
gen storage quantities, low overall system weight is an essential are arranged in a triangular configuration. Filters of radius ‘r3 ’
requirement. A well designed multiple tubes and filters config- are placed at the centre of each triangular arrangement. The
uration provides the possibility to optimize the system weight centre to centre distance between the cooling tubes is termed
for given specifications. In the light of the above, a parametric as the pitch distance ‘s’.
study of a hydrogen storage device with embedded filters and Various assumptions made for the simplified treatment of the
heat exchanger tubes highlighting the influence of important problem are given below.
geometric and operational parameters on the hydriding perfor-
mance of the device is presented in this paper. (i) Hydrogen is assumed as an ideal gas as the pressure within
the bed is moderate.
2. The physical model (ii) Effect of hysteresis and plateau slope is neglected in sim-
ulation as their corresponding values for LaNi5 is very
The metal hydride storage device considered in the present small [12].
study is shown in Fig. 1. The device contains several tubes and (iii) Natural convection and radiation effects within the bed
filters spaced uniformly inside a cylindrical shell. Hydrogen are neglected.
storage material evenly packed within the cylindrical storage (iv) The hydride bed is assumed to have uniform porosity and
vessel forms the reaction bed in which the cooling tubes and is isotropic. Brittle metal hydride materials upon cyclic
filters are located. Heat transfer fluid is circulated through the hydrogenation asymptotically reach a minimal particle
cooling tubes to carry away the reaction heat generated in the size [13]. Therefore, effect of powdering and variation in
sorption process. The filters serve the purpose of distributing porosity is more pronounced in the early stages of absorp-
hydrogen evenly within the bed. Reiterating the fact that ab- tion. After a certain number of charge–discharge cycles
sorption process will be faster in the close vicinity of cooling the variation in porosity with time is insignificantly small.
tubes and it will be saturated earlier than the rest, we can vi- (v) Pressure drop within the bed is negligible.
sualize a smooth interface, known as a reaction front which (vi) Local thermal equilibrium is assumed between the gas
demarcates the fully hydrided portion from the rest of the un- and solid within the bed. Non-local thermal equilibrium
saturated regions. As time proceeds, these reaction fronts move effects are confined to cooling tube wall and hydrogen
outwards from the periphery of cooling tubes to the unsaturated filters. However, its overall effect on the performance of
4980 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987

Fig. 1. Schematic diagram of cylindrical metal hydride reactor.

Table 1
Thermo-physical properties of LaNi5 and hydrogen [8,10,16]

Parameter LaNi5 Hydrogen

Density, kg m−3 8200 0.0838 at NTP

Specific heat, J kg−1 K −1 419 14,890
Effective thermal 1.2 0.12
conductivity,
W m−1 K −1
Diffusivity, m2 s−1 – 4.6 × 10−12
Permeability, m2 10−8 –
Activation energy, J mol−1 21179.6 –
Constants in Eqs. (2), (3): Ca 59.187 –
A 12.99 –
B 3704.59 –

Table 2
Parameter values used in the present simulation study

Parameter Range of values

Initial temperature of 293 (constant)

hydride bed, T0 (K)
Inlet pressure of hydrogen, pin (bar) 8.0–15.0
Cooling media temperature, Tf (K) 293–323
Fig. 2. Spatial variation of pressure and velocity of hydrogen within the metal
Porosity 0.5 (constant)
hydride bed after 3600 s absorption (b = 15 mm, r1 /s = 4 corresponding to
Overall heat transfer 500–1500
N = 55, d2 = 20 mm, Tf = 300 K, h = 500 W m−2 K−1 and p = 15 bar).
coefficient, h
(W m−2 K −1 )
Bed thickness, b (mm) 10–27.5
Cooling tube diameter, d2 (mm) 5–20 considered independent of bed temperature, pressure and
Ratio of container 3–7 (corresponding to N = 31.163) concentration.
radius to pitch (r1 /s) (viii) Isothermal condition is assumed at the cooling tube walls.

3. Problem formulation
the device is minimal [14]. In most bulk portions of the
bed, local thermal equilibrium is still valid. So it may be 3.1. Mass balance of metal
justified in simulation.
(vii) Thermo-physical properties such as thermal conductiv- Hydrogen entering the alloy bed through filters gets absorbed
ity, dynamic viscosity and permeability of the bed are and converted into metal hydride with associated changes in
 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987 4981

mol/m3

X 104

1.920

1.915

1.910

1.905

1.900

Fig. 3. Formation of hydride around cooling tubes during absorption at different time intervals (Tf = 300 K, p = 15 bar, h = 500 W m−2 K−1 , r1 /s = 4
corresponding to N = 55, b = 15 mm, d2 = 20 mm). (a) 600 s; (b) 1200 s; (c) 1800 s; (d) 2400 s; (e) 3000 s; (f) 3600 s.

density. Transient changes in hydride density are attributed to 3.2. Reaction kinetics
both absorption reaction and diffusive transport due to spatial
variation in concentration. As listed under assumptions, con- The amount of hydrogen absorbed is directly related with
stant values of porosity and diffusion coefficient are taken for the reaction rate, and could be given as follows [7]:
simulation. Thereby conservation of mass for the solid phase    
Ea P
of the reactor in polar coordinates is expressed in the following ṁ = −Ca exp − ln (sat − s ), (2)
form: RT Peq
where Ca is a material dependent constant, Ea is the activation
 
js 1 j js energy of the material, sat is the density of hydride at satura-
(1 − ) = ṁ + (1 − )D r tion. The change in volume if any, due to hydrogen absorption
jt r jr jr
  is not considered in the computation of density of the bed. The
1 j js
+ (1 − )D 2 , (1) equilibrium pressure is determined by the van’t Hoff relation-
r j j
ship as given below
B
where s denotes the density of the hydride, D denotes the dif- ln Peq = A − , (3)
fusion coefficient and ṁ represents the reaction term. Transient T
variation in density on the left-hand side is expressed as a sum where A and B are van’t Hoff constants.
of three terms on the right-hand side. The first term represents
the mass of hydrogen absorbed due to reaction. The second and 3.3. Darcy law
third terms of the equation signify the effect of diffusive trans-
port due to the concentration gradients existing within the bed. The following relation represents the flow of hydrogen in
Convection term is neglected in Eq. (1). porous bed. It combines the Darcy law and mass conservation
4982 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987

K
360

350

340

330

320

310

300

Fig. 4. Evolution of temperature in metal hydride bed at different time intervals (Tf = 300 K, p = 15 bar, h = 500 W m−2 K−1 , r1 /s = 4 corresponding to
N = 55, b = 15 mm and d2 = 20 mm). (a) 600 s; (b) 1200 s; (c) 1800 s; (d) 2400 s; (e) 3000 s; (f) 3600 s.

equation for hydrogen. Eq. (5) assumes the existence of local thermal equilibrium be-
   tween the solid and gas within the bed. Hence the heat trans-
jg − 1 j jp fer between the two phases is omitted. Radiation heat transfer
 + rg
jt  r jr jr to the ambient is also neglected. Heat of formation of hydride
   
− 1 j jp represents the source term in the equation.
+  = −ṁ. (4) Effective volumetric heat capacity is expressed as follows:
 r 2 j g j
(Cp )e = (g Cpg + (1 − )s Cps ). (6)
Darcy law considers the porous bed and the fluid as a single
homogenized media. The velocity vector is determined by the Effective thermal conductivity is expressed as given below
pressure gradient, the fluid viscosity and the structure of the
porous bed. ke = kg + (1 − )ks . (7)

3.4. Energy balance 3.5. Initial and boundary conditions

Heat generated due to exothermic reaction causes associated Initially the pressure and temperature of reactor bed are as-
spatial temperature imbalances within the bed. Energy conser- sumed to be uniform. Therefore,
vation equation of the bed could be expressed as given below
p = p0 ; T = T0 ;  = 0 at t = 0. (8)
   
jT 1 j jT 1 j jT
(Cp )e = ke r + ke 2 − ṁH 0 . (5) The boundary walls of the reactor are assumed to be imper-
jt r jr jr r j j meable and adiabatic. The reaction heat is removed from the
 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987 4983

mol/m3

X 104

1.920

1.915

1.910

1.905

1.900

Fig. 5. Formation of hydride around cooling tubes at 1800 s absorption for different bed thickness values (Tf = 300 K, p = 15 bar, h = 500 W m−2 K−1 ,
r1 /s = 4 corresponding to N = 55, d2 = 20 mm). (a) 5 mm; (b) 10 mm; (c) 12.5 mm; (d) 15 mm; (e) 20 mm; (f) 27.5 mm.

adiabatic. Based on these, the following boundary conditions

are written:
Reactor wall at r = r1 :
jp jT
= 0; =0 at t > 0. (9)
jr jr
Cooling tube wall at r = r2 :
jp jT
= 0; −k = hf (Tf − T ) at t > 0. (10)
jr jr
Filter wall at r = r3 :
jT
p = pin ; =0 at t > 0. (11)
jr
Fig. 6. Influence of bed thickness and elapsed time on absorption of hydrogen.
4. Simulation methodology

The simulation of the metal hydride bed employing the above

porous bed using heat transfer media circulated through the mentioned partial differential equations were carried out using
cooling tubes. “No slip condition” is valid at the reactor walls. COMSOL MULTIPHYSICS commercial code [15]. The model
Hydrogen is supplied using porous filter tubes. Convective flux is drawn using AUTOCAD software package and exported to
conditions prevail at the filter walls which are assumed to be COMSOL MULTIPHYSICS. An unstructured mesh of element
4984 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987

the mass transfer process in the hydride bed include the hydro-
gen gas pressure, hydrogen equilibrium pressure, hydride bed
temperature and the hydrogen concentration in the hydride. A
large amount of heat is generated due to the exothermic hydride
forming reaction. The reaction within the bed is proportional to
the pressure gradient between the gas supply pressure and equi-
librium pressure corresponding to the bed temperature. While
the supply pressure remains constant, the reaction is vigorous
when the bed temperature and the corresponding equilibrium
pressure are lower. The bed temperature increases rapidly dur-
ing the initial phase of the reaction. Heat transfer from the bed
to the cooling media dissipates the heat generated from the
process. Since the thermal diffusion process is comparatively
slower than the chemical reaction, the hydriding rate drops as
the time progresses.
Figs. 3 and 4 show the variation of concentration and temper-
ature with time for a typical case. Reaction rate is faster in the
vicinity of the heat exchanger tubes. Away from the tube wall,
the bed temperature rises rapidly to a peak value and decreases
slowly. Hydriding alloy requires large temperature gradients to
conduct the heat of reaction to the heat exchanger tubes due
to its low thermal conductivity. The high temperatures prevail-
ing farther from the cooling tubes increase the corresponding
Fig. 7. Schematic of different reactor configurations used in simulation with equilibrium pressure, thereby reducing the reaction rate. Hence,
different number of cooling tubes at constant bed thickness.
faster reaction rates in the vicinity of the tube lead to higher
saturation levels near the tube. As the reaction proceeds, the
size 0.001 m and finer is generated by defining suitable mesh reaction front progresses from the heat exchanger tube walls to
parameters. Most of the cases involve degrees of freedom in the the unsaturated areas within the bed.
range 50 000–100 000. The software package solves the mass
(convection–diffusion), momentum (Darcy’s law) and energy
equations (transient conduction) in differential form with spec- 5.1. Effects of system configuration
ified initial and boundary conditions. Suitable solver settings
ensure fast convergence of the problem. Grid independence Bed thickness is an important geometric parameter that in-
study is carried out to arrest the variations in results. fluences the performance of any hydrogen storage device. For
thinner beds, the conduction path is relatively short. If the thick-
5. Results and discussion ness exceeds an optimum value, the gain due to higher quantity
of hydride is negated by the increased resistance to heat and
Results are obtained for LaNi5 as the hydriding material. Fast mass transfer. Hence it results in a higher charging duration for
reaction kinetics, low temperature hydriding and its wide appli- the entire bed to get saturated. The above trend is illustrated
cability in storage applications makes it a material of choice in in Fig. 5, which shows the formation of hydride around cool-
the present study. Reaction kinetics and thermo-physical prop- ing tubes for different bed thicknesses. At specified conditions
erty data for LaNi5 are readily available in literature and are of cooling media temperature, charging pressure and number
given in Table 1. The simulation is carried out for various op- of cooling tubes, rate of absorption depends on bed thickness.
erating parameters, the ranges of which are given in Table 2. Fig. 6 shows the influence of bed thickness on absorption of
Performance of metal hydride device with embedded filters and hydrogen. In all cases, the same saturated state of about 1.2 w%
heat exchanger tubes demands the definition of two major geo- H2 is achieved after different time intervals. It is obvious that,
metric parameters namely, bed thickness, b and the ratio of the at lower bed thicknesses the hydride bed reaches saturation ear-
radius of the container to pitch (r1 /s). Once the bed thickness lier. Higher bed thicknesses follow a flatter profile and reach
is fixed to a constant value, r1 /s decides the number of cool- the saturation state much later. Higher bed temperature, lower
ing tubes, N within the bed for a specified system capacity and reaction rate, larger conduction path length and low thermal
configuration. conductivity of the material collectively contribute to the above
Distribution of pressure and flow within the bed is repre- behaviour. For a given bed thickness, both heat transfer and
sented in Fig. 2 for a typical case. Similar trend was seen over reaction kinetics are important in determining the absorption
the complete range of parameters. The spatial variation in pres- time. As reaction kinetics is more material dependent, for a
sure is negligible due to which the convection effects are of specified effective thermal conductivity value, there exists an
less significance. Diffusion plays dominant role in controlling optimum bed thickness satisfying the performance and system
the mass transfer in the bed. The main parameters which rule weight.
 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987 4985

mol/m3

X 104

1.920

1.915

1.910

1.905

1.900

Fig. 8. Formation of hydride around cooling tubes at 1800 s absorption for different values of r1 /s and corresponding numbers of cooling tubes (Tf = 300 K,
p = 15 bar, h = 500 W m−2 K−1 , b = 15 mm, d2 = 20 mm). (a) r1 /s = 3, N = 31; (b) r1 /s = 4, N = 55; (c) r1 /s = 5, N = 85; (d) r1 /s = 6, N = 121;
(e) r1 /s = 7, N = 163.

Fig. 9. Influence of number of cooling tubes and elapsed time on absorption

of hydrogen. Fig. 10. Effect of cooling tube diameters and elapsed time on hydride
formation.

Sizing of high capacity storage devices demand a judicious

selection of its diameter and length strictly based on various with less number of cooling tubes, or the one with more number
limiting considerations. For a prescribed storage capacity at of cooling tubes and shorter length. Temperature of the coolant
constant bed thickness, they either require a long storage bed increases as it flows through the bed dissipating the heat of
4986 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987

reaction. Long storage beds lead to lower performance levels

owing to higher temperature rise of the coolant. The above
effect is compounded by the axial pressure variation within the
bed. Devices with large diameters but shorter lengths overcome
the above drawbacks and also ensure an even distribution of
hydrogen within the bed. However, they could turn out to be
bulky.
Fig. 7 depicts arrangements of different number of cooling
tubes in the storage bed. The configuration follows the scheme
described earlier in the physical model. The ratio, ‘r1 /s’ as-
sumes different values ranging from 3 to 7. Pitch retains a con-
stant value as the bed thickness and cooling tube radii are con-
stant. Keeping r1 /s = 3, 4, 5, 6 and 7 correspondingly require
31, 55, 85, 121 and 163 tubes to be arranged in the specified
manner. Fig. 11. Effect of supply pressure on hydride formation.
As depicted in Fig. 4, hotspots are observed in the periph-
eral regions of the bed, which get saturated much later than the
rest of the device. The extents of such regions are analogically
similar to the void spaces of a cylindrical shell stacked with in-
dividual cylinders of identical diameters. Bed thickness and the
ratio of container radius to pitch (r1 /s) collectively determine
the existence of such unsaturated regions, which are quantita-
tively much smaller compared with the rest of the device. For-
mation of hydride at different values of ‘r1 /s’ after 1800 s is
illustrated in Fig. 8. Disparity among different cases is evident
only in the peripheral regions and the inner regions develop a
uniform growth of saturated fronts. Based on the above fact,
it may be inferred that, bed thickness is rate controlling at all
spatial locations within the bed and throughout the hydriding
process. Effect of ‘r1 /s’ is limited to the peripheral regions and
towards the closing stages of reaction. However, there exists Fig. 12. Effect of cooling media temperature on hydride formation.
an optimum ratio corresponding to a fixed number of cooling
tubes which dissipates heat better than the rest at all different
locations within the bed and thereby minimizes the unsaturated
space. The above statement explains the apparent gain at N =85
(r1 /s = 5) over other cases, towards the final stages of reaction
(t > 3000 s) as seen in Fig. 9.
Fig. 10 shows the influence of tube size on hydriding time.
The analysis assumes isothermal conditions and constant heat
transfer coefficient at the walls. Bed thickness is also kept
constant. Increasing the heat exchanger tube diameter leads to
higher heat transfer surface area per unit available volume of the
container, which causes higher heat transfer rates and thereby
better hydriding performance. However, increasing the heat ex-
changer tube diameter correspondingly increases the size of the
container as well as the percentage of unsaturated area within
the whole available area of the bed, which acts as a negat- Fig. 13. Influence of tube side heat transfer coefficient on hydride formation.
ing factor. So the gain in hydriding performance progressively
comes down as we increase the cooling tube diameter. Selection operational parameters which influence the performance of the
of tube diameter should also be determined by the hydraulic hydride bed.
losses within the flow passage. Supply pressure is an important operating parameter con-
trolling the hydriding rates because the differential between
equilibrium pressure and imposed pressure within the reactor
5.2. Effects of operating conditions controls the absorption process. A higher pressure differential
reduces the elapsed time for hydriding as seen in Fig. 11. Two
Hydrogen supply pressure, cooling media temperature and other parameters namely, cooling media temperature and cool-
cooling tube side heat transfer coefficient are independent ing tube side heat transfer coefficient control the transport of
 G. Mohan et al. / International Journal of Hydrogen Energy 32 (2007) 4978 – 4987 4987

exothermic heat generated within the bed. Effect of cooling me- [2] Yu WS, Suuberg W, Waide C. Modeling studies of fixed bed metal
dia temperature on hydriding is shown in Fig. 12. At any given hydride storage systems. Hydrogen energy progress—Part A, Proceedings
supply pressure, lower coolant temperature causes a lower bed of THEME Conference, Florida, 1974. New York: Plenum Press; 1975.
p. 803–17.
temperature and thereby lower equilibrium pressure. It conse- [3] Goudy AJ, Stokes DG, Gazzillo JA. The effect of heat transfer on the
quently increases the driving potential for mass transfer, re- desorption kinetics of LaNi5 H6 . J Less-Common Met 1982;91:149–58.
sulting in a higher rate of reaction. Impact of tube side heat [4] Kawamura M, Ono S, Higano S. Experimental studies on the behaviour
transfer coefficient on hydrogen absorption is small as observed of hydride heat storage systems. Energy Conver Manage 1982;22:
in Fig. 13. 95–102.
[5] Ram Gopal M, Srinivasa Murthy S. Studies on heat and mass transfer
in metal hydride beds. Int J Hydrogen Energy 1995;20:911–7.
6. Conclusion [6] Nasrallah SB, Jemni A. Heat and mass transfer models in metal-
hydrogen reactor. Int J Hydrogen Energy 1997;22:67–76.
Simulation of a metal hydride hydrogen storage device with [7] Mat MD, Kaplan Y. Numerical study of hydrogen absorption in Lm.Ni5
embedded filters and cooling tubes is carried out with LaNi5 hydride reactor. Int J Hydrogen Energy 2001;26:957–63.
[8] Dogan A, Kaplan Y, Veziroglu TN. Numerical investigation of heat and
as the alloy material for various parametric values. Results of mass transfer in a metal hydride bed. Appl Math Comput 2004;150:
the simulation reiterate the importance of bed thickness as a 169–80.
major parameter controlling the rate of hydrogen absorption at [9] Askri F, Jemni A, Nasrallah SB. Study of two-dimensional and dynamic
all locations within the bed. Effect of the ratio of container ra- heat and mass transfer in a metal-hydrogen reactor. Int J Hydrogen
dius to pitch (r1 /s) on hydrogen sorption cannot be ignored. Energy 2003;28:537–57.
[10] Askri F, Jemni A, Nasrallah SB. Prediction of transient heat and mass
Heat exchanger tube diameters are observed to have marginal transfer in a closed metal-hydrogen reactor. Int J Hydrogen Energy
influence on hydriding time at higher values. Design of hy- 2004;29:195–208.
drogen storage device with embedded filters and cooling tubes [11] Demircan A, Demiralp M, Kaplan Y, Veziroglu TN. Experimental and
demands a suitable method for the calculation of above geo- theoretical analysis of hydrogen absorption in LaNi5 .H2 reactors. Int J
metric parameters satisfying the prescribed performance and Hydrogen Energy 2005;30:1437–46.
[12] The hydride information centre containing thermodynamic and
system weight. Studies of parameters which include hydrogen hydrogen storage properties of selected metal hydride alloys,
supply pressure, coolant temperature and tube side heat trans- http://www.hydpark.ca.sandia.gov.
fer coefficient highlight their influence at different operating [13] Ron M. The normalized pressure dependence method for the evaluation
conditions. of kinetic rates of metal hydride formation/decomposition. J Alloys
Compd 1999;283:178–91.
[14] Lloyd GM, Razani A, Kim KJ. Formulation and numerical solution of
References non-local thermal equilibrium equations for multiple gas/solid porous
metal hydride reactors. Trans ASME-J Heat Transfer 2001;123:520–6.
[1] Powers GJ, Cummings DL. The dynamics of energy and mass transfer [15] Comsol Multiphysics User’s Guide. Comsol AB, Sweden; 2005.
with reaction in hydride storage system. Hydrogen energy progress—Part [16] Majer G, Kaess U, Bowman Jr. RC. Nuclear magnetic resonance studies
B, Proceedings of THEME Conference, Florida, 1974. New York: Plenum of hydrogen diffusion in LaNi5.0 H6.0 and LaNi4.8 Sn0.2 H5.8 . Phys Rev
Press; 1975, p. 803–17. B 1998;57:599–603.